Niebyl 1980.

Methods	Randomised placebo‐controlled trial.
Participants	32 women between 24‐33 weeks' gestation with preterm labour (defined as cervical dilatation of ≥ 2 cm or 75% effaced or any cervical change during a period of observation together with painful/regular contractions). Exclusion criteria: ruptured membranes, cervical dilatation > 4 cm, FGR, salicylate sensitivity or peptic ulcer disease history, intrauterine infection and IUGR.
Interventions	COX group: indomethacin. Bolus dose 50 mg (p.o.) followed by 25 mg every 4 h up to 24 h. If contractions recurred after discontinuation, the protocol was repeated with the same study drug for another 24 h. Control group: placebo. Lactose capsule (p.o.). Both groups: alternative therapy was used for treatment failure (defined as cervical dilatation > 4 cm > 2 h after initial dose).
Outcomes	Maternal: mean plasma concentration of treatment drug, gestational age. Fetal/neonatal: Side effects. Benign: Bleeding, NEC, respiratory distress, hyperbilirubinaemia, metabolic abnormality. Life‐threatening: pneumothorax, BPD, severe RDS, persistent fetal circulation, PDA with congestive heart failure, bradycardia and premature atrial contractions, chronic hepatitis, hydrocephalus. Causes of death: NEC, sepsis, stillbirth.
Notes
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Capsules were placed in envelopes and allocated to patients according to a table of random numbers..."
Allocation concealment (selection bias)	Unclear risk	"Capsules were placed in envelopes..." The authors consider this approach unclear risk.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	"Capsules were allocated to patients [...] in a double‐blind fashion." The authors consider this approach low risk.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Based on blinding of participants and personnel, the authors consider this low risk.
Incomplete outcome data (attrition bias) All outcomes	Low risk	2 women were excluded from the indomethacin group: 1 woman vomited the initial capsule treatment and no detectable levels of indomethacin were achieved and 1 woman was treated in error.
Selective reporting (reporting bias)	Low risk	All outcomes were reported as expected.
Other bias	Low risk	None apparent.